FIELD THERMAL PERFORMANCE OF RADIANT BARRIERS AND INTERIOR RADIATION CONTROL COATINGS FOR ATTIC RETROFITS

Ali Fallahi1, David Yarbrough

2, Jan Kośny

3

ABSTRACT For over a century, traditional bulk insulation materials like fiberglass and cellulose have been

used in attics to prevent heat from escaping or entering American homes. But times are changing

and today, builders and homeowners in Southern U.S. locations are discovering that adding attic

reflective insulation may offer significant gains in thermal efficiency – often far greater than the

same investment in additional layers of conventional bulk insulation. According to the US

Information Administration (EIA4), in 2009 in average more than 6% of the US household end-

use energy expenditure was for space cooling (air conditioning) while in the US hot humid

climate region this space cooling expenditure went up to more than 25%. In order to reduce

cooling energy consumption, techniques to limit radiative heat transfer such as Radiant Barriers

(RB) and Interior Radiation Control Coatings (IRCC) can be applied to existing residential attics

in single family homes or light commercial buildings. RBs incorporate two layers of aluminum

foil or aluminized plastic film. Aluminum has a low emissivity, absorbing and emitting a small

amount of infrared radiation. IRCC work in a similar manner, but it is a coating usually sprayed

on the back side of the roof deck. There are claims that RB and IRCC are easier to install and

very energy efficient in retrofit projects. The goal of this study is to understand these claims with

analyzing changes in the annual cooling energy consumption and peak cooling loads by both

experimental and numerical approaches. In Austin, Texas, both RB and IRCC were installed in

test houses and compared to a baseline house with no modifications. The test houses were

instrumented and thermal and energy performances of the attics were monitored for over 6

months. In addition the whole building energy consumption before and after retrofit was

compared using EnergyPlus energy consumption simulations. 1. Introduction Attic RBs and IRCCs present a unique way of increasing the thermal performance of existing or

new insulation within the space between the roof deck and ceiling level in residential and small

commercial buildings. Attic RBs save energy by reducing the transfer of heat from the hot roof

to the attic floor insulation in the summer. The transfer of heat from the attic floor insulation to

the roof in the winter is also reduced.

Aluminum surfaces, like those found in RB applications, have thermal emittances in the range of

0.03 to 0.06. Therefore, there is very little radiant transfer across a space bounded by a RB. Most

often, RBs are aluminum foil laminates or aluminized synthetic films sheets. The foils are

laminated to paper, most commonly to Kraft paper, synthetic films, oriented strand board (OSB),

1 Ali Fallahi, Ph.D., R&D staff, Building Science Engineer, Fraunhofer CSE, Boston, MA

2 David Yarbrough, Ph.D., P.E., President, R&D Services, Cookeville, TN

3 Jan Kośny, Ph.D., Building Enclosure Program Lead, Fraunhofer CSE, Boston, MA

4 http://www.eia.gov/consumption/residential/data/2009/index.cfm?view=consumption#end-use

or plywood. For the aluminized synthetic films, a thin layer of aluminum particles are deposited

on the films through a vacuum process. These laminates and films are characterized by having at

least one low-e surface of 0.1 or less (ASTM C 1313, 2010).

IRCCs are low-e coatings or paints that, when applied (i.e., sprayed or painted) to building

surface (e.g., OSB, plywood, metal siding, or plasterboard), decrease the emittance of these

surfaces to 0.25 or less (ASTM C 1321, 2009). Both RBs and IRCCs have received considerable

attention due to their potential to reduce the radiant heat transfer across vented spaces between

roofs and ceilings of buildings (e.g., attic spaces in residential buildings). In the case of RBs,

aluminum is used because it is inexpensive and is a surface that, once exposed to air, becomes

covered with a layer of a transparent oxide that protects it from the atmosphere and allows it to

maintain a constant emittance for long periods of time5.

In some cases, an RB can include an enclosed air space in order to provide thermal resistance to

the path between the roof sheathing and the attic floor. In this study an RB was used with an

enclosed air space, which is a common configuration with multi-layer products. In some cases,

there is an R-value6 7 associated with the reflective insulation component application. In the

configuration considered in this research, the thermal performance of interior-facing RB has to

be combined with the low thermal conductive performance of the enclosed air cavities.

Low-e coatings (IRCCs) have similar heat reduction principles as RBs in residential attics.

IRCCs are also being studied in this work. This is a crucial study for the industry since there are

very few examples of directly measured savings due to spray-applied or liquid-applied attic

IRCCs.

Base on Medina (2012), experimental results highlighting cooling load and ceiling heat flux

reductions produced by RBs are vary depending on nominal ceiling insulation R-value, testing

protocol, climate zone, ventilation type, occupancy and duct inclusion in the attics. Medina

(2012) reported that in attics with nominal insulation levels of R-11, R-19 and R-30, average

space cooling reductions are 14%, 20% and 6%, respectively. He also reported the average

reduction in heat flow produced by installing RBs in attics with insulation levels of R-11, R-19

and R-30 are 45%, 30% and 23%.

According to Medina (2012), laboratory-controlled experiments of IRCC applied in a flat system

configuration with an insulation level of R-19, produced average heat flow reductions of 32%

(vs. the same system without the application of any coatings).

Although the industry does not have software to evaluate energy savings, there is an ASTM

5 http://www.lenntech.com/periodic/elements/al.htm

6 Thermal resistances calculated for an enclosed air space with one surface having emittance 0.03 and a second

surface with emittance 0.9 are 1.99 and 3.43 (ft2•h•°F/Btu) when heat directions are up at 45º and down at 45º,

respectively. The thermal resistance values in the report do not include the effect of the air space between the

product and the roof sheathing of the RB effect due to a low-emittance surface facing the attic floor.

7 http://www.rimainternational.org/index.php/technical/library/residential-commercial/

consensus method for calculating RB performance (ASTM C 1340, “Standard Practice for

Estimation of Heat Gain or Loss through Ceilings under Attics Containing Radiant Barriers by

Use of a Computer Program”). This program is called “AtticSim” and was developed by

Kenneth Wilkes. In the past, it has been used by Oak Ridge National Laboratory to generate

input for two editions of DOE RB fact sheets.

Currently, RBs and IRCCs are difficult to numerically analyze using whole building energy

simulation tools. In the past, DOE 2.2, EnergyPlus, and BEopt, have not had the capability for

detailed attic modeling. Recently, there has been an effort to link C 1340 to DOE 2.2 and provide

modeling of attic RBs. However, the status of this project is unclear. Current DOE-sponsored

work with AtticSim is limited to attic RBs and does not extend to evaluations of reflective

insulation assemblies.

In the current study in collaboration with a team of local companies, both RB and IRCC

radiation control technologies were tested in the field condition. One type of RB product with

enclosed air space and two types of IRCC products were tested. For this puporse, four test

houses (two duplexes) in Austin, TX were chosen. These radiation control systems were installed

in the attic of three test houses.One attic, where none of the radiation control technologies were

installed, established the baseline.

All attics had heat flux transducers and thermistors installed to monitor both surface and air

temperatures. Comparisons of attic air temperatures were made to assess system performance.

Although heat flux is an indicator of the heating or cooling loads, air temperature reflects the

unique conditions of each unit as well as the loads. These unique conditions included varying

parameters within the four units that could not be changed or controlled. Notable variations

between the four cases include:

Ventilation of conditioned space– Occupants operating windows and adjusting HVAC

systems varied the flow rate greatly

Air leakage through ceiling– Craftsmanship of ceiling construction varied by apartment

(i.e. holes around pipes, ducts, etc.)

R-value of attic– Construction and insulation type and thickness varied between duplexes

Internal temperatures– Occupants adjusted the thermostat causing non-uniform

temperatures between units.

Finally, representative computer models were constructed to further enhance understanding of

the radiation control systems. Comparisons to the utility consumption data were made from the

simulation results.

2. Field Testing The test houses consisted of two duplex houses in Austin, which were closely located to each

other (< 0.1 miles apart). The duplexes were both constructed of 2×4 wood framing with brick or

wood cladding. Duplex 1 was built between 1987 and 1997 and Duplex 2 was a pre-1987

construction. Each duplex has two attached, unconditioned garages. The windows were operable,

single glazed in non-thermally-broken aluminum frames.

The attics over the garages are separated from the attics over the houses and were not installed

with the radiant control technologies. In order to thermally separate the test units, the attics over

the houses were divided by insulated walls for the purpose of this field study. The ceilings were

flat in Duplex 1 and vaulted in Duplex 2. Overall, the insulation was very non-uniform in all test

attics. In several places, the loose-fill cellulose could reach depths such that the R-values equal

approximately R-30 but were lower at other places. Four different test cases were established

(See Figure 1):

1. Duplex 1

A. Nothing added to the east attic (used as baseline)

B. IRCC (TYPE I) added to the west attic

2. Duplex 2

A. RB added to the east attic

B. IRCC (Type II) added to the west attic

There was no opportunity to measure the performance of the four units prior to the retrofits in

order to establish a baseline in each unit without radiation control technologies. There was an

option to make the aforementioned insulation levels consistent; however, this would have made

the utility bill calibration comparison impossible. For this reason, it was decided to maintain

post-retrofit conditions, such as infiltration and insulation, as closely as possible to the pre-

retrofit conditions. The only exception to this was the addition of rigid polyisocyanurate

insulation in the attic as a vertical divider to thermally separate the two units. Considering the

attic air temperature difference created by applying these technologies is small, the applied

insulating divider isolates the effect of each technology from each other.

Figure 1: Test Houses in Austin, TX - Duplex 1 (top); Duplex 2 (bottom). Location of thermistor arrays and heat flow transducers inside the tested attics

2.1 Instrumentation Each partitioned attic has been instrumented to measure thermal energy flows and included

thermistor arrays for the test attics and heat flux transducers on the ceiling between the living

zone and the attic zone. In addition, a weather station was installed to measure outdoor climatic

parameters. The attic instrumentation diagram is illustrated in Figure 2 for the two different

installed technologies.

Temperature was measured using shielded thermistors. Two arrays of thermistors were installed

at each attic: one on the underside of the south-facing roof and the other under the north-facing

roof. Heat flux transducers measured the energy flow through the attic floors. The heat flux

transducers before deploying to the test houses were first calibrated between two pieces of

gypsum in a heat flow meter.

Figure 2: Diagrams of the attic sensor distributions for a RB combined with an enclosed reflective air space (right diagram), and IRCC and baseline (left diagram).

Figure 3: Installation of IRCC (left) & RB with enclosed airspace underside of the roof deck (middle, right)

2.2 Radiation Control Technology Installations As stated earlier, two types of low-cost radiation control strategies for southern U.S. applications

were installed in tested attics (Figure 3):

a) RB combined with an enclosed reflective air space installed in Duplex 2A. The installed RB in

this study is an insulating product composed of multiple layers of low-e materials designed to

reduce radiant heat transfer. The inside layer is a metalized polymer with emissivity of 0.04 and

the outside layer is reinforced aluminum foil kraft paper with emissivity of 0.03 bonded with a

fire-retardant adhesive. The layers expand when installed to form a reflective air space to provide

enhanced thermal performance and protect the low-e surface from the performance reducing

effects of dust accumulation.

b) spray-applied IRCC. Two types of IRCCs were sprayed: IRCC Type I- This was applied to

the attic of Duplex 1B and it is capable of lowering their surface emissivity to 0.19 or lower.

IRCC Type II- It was sprayed to the attic of Duplex 2B and lowers surface emissivity to 0.17 or

lower.

3. Data Analysis In the following sections, the recorded test data collected during June/July 2012 is presented and

compared for all four test units. Note that all test housing units were occupied during the testing

by families. Each of these families had different occupation habits and thermal comfort

preferences. In each house, thermostat temperature setups were close, but not identical. Also,

individual schedules for using the AC were different. In addition, the duplexes—even though

they had been similarly built in relatively close time periods—had attics with different structural

components, thicknesses, types of attic floor insulation and levels of attic ventilation. Significant

attention was paid to the comparison of the attic air temperatures, which were considered (due to

the listed above differences in the attic construction) as a best indicator of the performance

differences between analyzed technologies.

The following data are summarized and analyzed below:

Internal Air Temperature

Roof Surface Temperature

Attic Air Temperature

Attic Insulation Surface Temperature

Ceiling Heat Flux

3.1 Internal Air Temperature Comparisons Figure 4 (Left) displays that the internal temperatures were close to 25°C in June. Most likely,

residents of Units 1A and 2A preferred interior temperatures slightly above 25°C, while residents

from Units 1B and 2B preferred temperatures closer to 24°C.

Figure 4: Average interior air temperatures recorded during the June/July measure period (Left); Maximum roof surface temperatures recorded during the June/July measure period (Right).

3.2 Attic Surface Temperature Comparison Figure 5(Right) shows daily average surface temperature of the attic surfaces recorded during

June/July. When the weather was warm in June, the roof temperatures of Duplex 2 were greater

than Duplex 1. The maximum roof temperatures in June for Duplex 2 were up to 5°C hotter than

Duplex 1. This was most likely due to general differences in construction of both types of attics

and thermal behavior of the RB combined with the enclosed air gaps, installed under the roof

deck in the test house 2A.

3.3 Attic Air Temperature Comparisons Typically in residential attic applications, RBs and IRCCs work by reducing the amount of

thermal radiation that is transferred across the air space between the roof deck and the top of the

attic floor insulation. Since the amount of thermal radiation increases as the temperature of the

emitting surface increases, it is critical to keep it as low as possible. This surface can be directly

roof deck, foil-laminated roof deck boards, roof deck boards coated with IRCC, or RB material

facing the interior of the attic.

An application of multi-foil RB in the case of Unit 2A showed that it was possible to apply

radiation control technology, increase the roof deck temperature and at the same time reduce the

air temperature in the attic. As depicted in Figure 4 (Right), this technology generated a slightly

higher roof surface temperature compared to other test attics. However, when attic air

temperatures were compared, the recorded attic temperature for Unit 2A was significantly lower

comparing to other attics; see Figure 5(Left) and Figure 5(Right).

Figure 5: Average daily attic air temperatures recorded in four test attics (center location) during the June/July measure period (Left); Maximum daily attic air temperatures recorded in four test attics (center location) during the June/July measure period (Right)

In this study, due to the structural differences between the test attics and due to the fact that the

attic floor insulations were not uniform in all the units, measurements of the attic air

temperatures in the center-of-the-attic location were used as one of indirect indicators of the attic

system thermal performance. The second potential performance indicator considered in this

project was heat flux measured on the ceiling level. However, as shown in Figure 4(Left), the

internal air temperatures were not identical in all test units. That is why direct comparisons of

measured heat fluxes cannot be used as a full indicator of the attic system thermal performance.

Figure 5 shows that the attic of Duplex 1 was typically hotter than Duplex 2 in June 2012. This is

consistent with the observation that the roof of Duplex 1 was cooler than Duplex 2. It is good to

remember that there were some differences between both duplexes, such as the roof deck

construction—rafters versus trusses—and the levels of ventilation in the attics, determined by

measuring the area of the soffit vents. An average baseline attic temperature for the considered

time period was 35.7°C (Figure 5 Left), while for the coolest attic 2A was only 33.4°C. This was

a 2.3°C difference.

Figure 5 (Right) depicts the maximum daily attic air temperatures for June/July 2012. These

temperatures reflect differences in attic-generated peak-hour cooling loads during the day. It is

clear that different types of reflective insulation work effectively in southern U.S. climates. In

June, the base case attic, 1A, had consistently higher maximum daily attic air temperatures. Units

1B and 2B were quite similar, but were generally a lower maximum temperature than the base

case (over 3°C on average). The fact that 1B and 2B appeared similar was a different pattern than

previously observed. However, examining the hotter time period in July may confirm that 2B

generally had lower attic temperatures than 1B. Finally, 2A, the house with the foil technology

installed, had the lowest maximum temperatures across this period by approximately 7°C, while

compared to the baseline attic, 1A.

As mentioned earlier, field measurements of the attic air temperatures were used in this study, as

one of the indirect indicators of the attic system thermal performance. It can be said that the attic

air temperature represented thermal condition equilibrium incorporating radiation, ventilation,

and convection effects together. This temperature also reflected structural differences between all

test attics.

The chart in Figure 6 (Left) indicates that the foil technology was thermally more effective than

both IRCC technologies. An average temperature difference between this attic and the base case

was about 3.4°C, while for the IRCC technologies it was about 2.4°C. In addition, the IRCC in

2B was usually more different from the base case than the 1B case.

Figure 6: Differences in average daily attic air temperatures recorded between three test attics using radiation control technologies and a baseline attic, during the June/July measure period (Left); Differences in maximum daily attic air temperatures recorded between three test attics using radiation control technologies and a baseline attic, during the June/July measure period (Right)

It is important to note that 3.4°C reduction of the average attic air temperature caused by the

multi-foil RB can be an equivalent to more than 34% reduction of the attic-generated cooling

loads—assuming approximate steady-state heat transfer, average attic temperature of 35°C and

internal air temperature of 25°C. It has been reported that for U.S. residential houses, roofs and

attics generate an average of 12% cooling energy contributions (Huang et al., 1996). In that light,

potential whole building cooling energy savings may reach about 4%, depending on other

building parameters and HVAC system efficiency.

Similarly, a 2.4°C reduction of the average attic air temperature due to the IRCC technology can

be an equivalent to more than 24% reduction of the attic-generated cooling loads. Potential

whole building cooling energy savings for this technology may be close to 3%.

Figure 6 (Right) shows clear superiority of the multi-foil radiation control technology used in the

attic 2A. For this attic, maximum air temperatures were, on average, about 8.5°C lower than the

base case attic. It can be also observed that IRCC systems reduced maximum attic air

temperature by over 6°C.

Following earlier analyses performed for the average attic temperature, a 8.5°C reduction of the

maximum attic air temperature caused by the multi-foil RB can be an equivalent to over 30%

reduction of the attic-generated peak-hour cooling loads, considering maximum attic temperature

of 53°C (as in Figure 5-Right) and internal air temperature of 25°C. Attic air temperature

recorded in Unit 2A was the most different from the base case. Given the hotter temperatures in

June, Unit 2B was typically the next most different attic temperature, as seen in Figure 6-Right.

Figure 7: Maximum daily temperatures on the top surface of the attic insulation recorded in three test attics using radiation control technologies and a conventional base case, during the June/July measure period (Left); Heat fluxes recorded in test attics using radiation control technologies and a conventional base case, during the June/July measure period. Positive heat flow direction is from conditioned space to the attic (Right).

In ideal conditions (when measurement perimeters were highly controlled and the insulation

layer is uniform), temperature measurements on the top of the attic insulation could be a very

good direct indicator of the technology thermal performance. However, in this experiment, due

to the differences in construction, thicknesses of insulation and types of insulation in the

individual attics, direct thermal performance comparisons were not possible using top insulation

surface temperatures.

3.4 Attic Insulation Surface Temperature Comparisons In the hot month of June, the insulation surface temperatures were still quite similar. Unit 2A

appears to have less extreme temperatures, as shown in Figure 7 (Left). Maximum day

temperatures on top of the attic insulation in Unit 2A were about 4°C to 5°C lower comparing to

the conventional base case attic. As mentioned earlier, these temperatures reflected the attic-

generated peak hour cooling loads.

3.5 Ceiling Heat Flux Comparison Because of different attic configurations and internal air temperatures, the heat fluxes measured

on the ceiling level were not good indicators of performance compared to attic temperatures.

There were four factors that can affect this measurement: ventilation rates, R-values, internal

temperatures, and air leakage through the attic ceiling. In June, the measured heat flux in the

house with IRCC technology was significantly higher than in the house with the foil technology,

as seen in Figure 7 (Right). The largest differences are between units 2A and 1B. Heat fluxes

measured in unit 2A were between 40% and 50% lower than heat fluxes measured for unit 1B.

4. Energy Modeling

4.1 Model Description The two duplexes in Austin were modeled using EnergyPlus to predict potential energy savings

as a result of applying RB and IRCCs technologies. The houses were each modeled as three

zones representing the living space, attic, and garage (a total of six zones in each duplex).

Outside boundary conditions came from a weather file for Austin’s climate. Wall and ceiling

constructions were modeled in THERM to determine the effective thermal conductivity of the

assembly layer containing the wood studs and insulation. As stated earlier, Duplex 1 was built

between 1987–1997 and Duplex 2 was built prior to 1987. Compared to parameters listed in

ASHRAE Handbook of Fundamentals Chapter 15, operable single-glazed windows with non-

thermally broken aluminum frames, such as those installed in the duplexes, were confirmed.

Solar heat gain coefficients were set to 0.79 in the model and U-factors were 0.92 Btu/h·ft2·°F

(5.2 W/m2·K). Window construction was also kept consistent throughout the models.

Roofs were modeled as shingles and plywood with addition of the appropriate paint or foil layer

where applicable. A no-mass material was specified in EnergyPlus to represent the IRCC and RB

technologies. The absorptance of the material was lowered to decrease the amount of radiant

transfer to other surfaces in the attic zone. Because of the non-uniform levels of insulation on the

attic floor, an average value was calculated for use in the model. It was determined that both

duplexes had approximate R-20, despite having different compositions—7.5 inches of loose

cellulose in Duplex 1 while 3.5 inches of batt insulation plus an average of 3.5 inches of loose

cellulose in Duplex 2, both with framing factors of 25%. Attic ventilation rates were estimated

based on measurements of the area of attic vents in each of the test houses (Atherton, 2011).

Cooling setpoints were kept for modeling purposes at 76.1°F (24°C) for all four units. The

heating setpoint was 68°F (20°C). Considering stratification of the temperatures, these

temperatures were very close to that which was measured in 2012 in the test houses. However, it

is important to notice that due to a change of some tenants followed by a possible change of

space conditioning preferences, measured temperatures in two units were about 1°C higher from

the other two. Two HVAC systems were added to each building, one per conditioned zone in

each duplex. The Unitary Template in EnergyPlus was used to model the forced air system with

a 2.5 rated coefficient of performance (COP) single-speed direct expansion cooling coil and

natural gas heating coil with an efficiency of 70%. Considering the losses due to duct leakage

and the condenser unit, the overall COP of the system was reduced in the model from the

manufacturer’s COP. The natural gas water heater had a thermal efficiency of 80%. The models

were calibrated against gas consumption data obtained from utility bills during winter. The

maximum hot water flow was obtained through calibration of gas consumption with the amount

reported by Texas Gas Service calculator for a typical residential single family house in Austin.

4.2 Calibration of Whole Building Energy Model In order to predict energy consumption of the duplexes with reliable accuracy, the EnergyPlus

models were calibrated against historical utility bills obtained for tested houses. For this purpose,

the computer-generated energy consumptions were compared with historical energy bills for

both heating and cooling seasons.

It is important to mention that, due to complexity of the test attics and the fact that buildings

were occupied during the testing, it was impossible to validate the thermal simulation algorithm

used for attics by EnergyPlus. Therefore, EnergyPlus simulation results (cooling energy savings)

presented in this study need to be confirmed in the future either by more accurate computer

models (like ATTICSIM–ASTM C1340) or by calorimetric field measurements with use of the

test huts. Additionally, due to limited amount of available historical bill data, the comparison

was done only for duplex 1B.

During the heating season, the main sources of energy consumption were a furnace for space

heating and a water heater for domestic hot water. Based on Texas Gas Service’s home energy

calculator, for a typical single family house with similar characteristics to the test houses, close

to 50% of total gas consumption was space heating, approximately 35% water heating and 15%

was cooking. Therefore, to capture these main sources of gas consumption, a water heater,

furnace and gas equipment were considered in the model. Figure 8 compares the simulated gas

consumption with historical energy bill and Texas Gas Service for duplex 1B. There was a

relatively good agreement between EnergyPlus-generated gas consumption and historical gas

service data. Note that EnergyPlus used the actual weather file which is the observed weather

data for the duration of the modeling. Figure 8 compare gas consumption of duplex 1B with

historical gas bill before and after applying the radiation technology. The data confirms proper

selection of the building enclosure and HVAC system parameters for whole building energy

consumption simulations.

Using the actual weather data file enabled comparisons to energy consumption results generated

with a use of the Texas Gas Service’s home energy calculator8. During the cooling season, the

main sources of electricity consumption were AC unit for space cooling, lighting and appliances.

Again, referring to Texas Gas Service’s home energy calculator, for a typical single family

8 http://www.texasgasservice.com/SaveEnergyAndMoney/HomeEnergyCalculator.aspx.

house, close to 80% of total electricity consumption was for space cooling, 8.5% for lighting and

close to 11.5% for appliances and others.

Figure 8: Comparison of monthly historical gas bill, simulated consumption, and the gas consumption by Texas Gas Service calculator. Gas consumptions are for duplex 1B before (Left) and after (Right) installing radiation control technology. The EnergyPlus simulation is based on actual weather file data from 2009 (before retrofit) and 2011 (after retrofit).

Figure 9: Comparison of monthly historical electricity bill, simulated consumption, and the electricity consumption by Texas Gas Service calculator. Electricity consumption is for Duplex 1B before (Left) and after (Right) installing radiation control technology. The EnergyPlus simulation is based on actual weather data file from 2009 (before retrofit) and 2011 (after retrofit). High electricity consumption during heating season is due to unexpected usage of the electric heater by the tenants. A cooling system with COP=2.5 was considered throughout this study; however, as a parametric study, cooling system with COP=2 was modeled and compared in this graph.

As Figure 9 shows, EnergyPlus predicted well the electricity consumption during the cooling

season compared with historical electricity bill and Texas Gas Service’s calculator. However,

during the heating season, there were discrepancies due to the fact that the test houses were using

auxiliary electricity for heating. Overall, considering the fact that the historical electricity bill

was not prorated and both the AC unit set point and consumer behavior were approximated, the

prediction of energy consumption particularly during the cooling season seemed in good

agreement with historical bills and Texas Gas Service’s calculator.

4.3 Prediction of the Cooling Energy Consumption As described, the models’ calibration was based on comparison between simulated energy use

and the electricity and gas consumption data and also with historical data from the Texas Gas

Service calculator. The calibrated models were simulated with a TMY3 weather file for Austin

Mueller Municipal Airport to predict the electricity consumed by AC units during cooling

season.

4.4 Modeling Results In the summer, a higher infiltration rate of the attic reduced the attic temperature by exchanging

the outdoor air with the attic air that was heated by the roof. The lower attic temperature reduced

the heat flux through the ceiling of the living space. The wall construction and infiltration rate of

the living space also affected the cooling load placed on the HVAC system. Fine tuning the

IRCC and RB was required to model the emissivity within the attic space accurately and the

thermal conductivity through the layer, and in the case of the RB, the air gap between the roof

and the RB.

The results from EnergyPlus using a typical meteorological year (TMY3) indicate that the

energy consumption is well below the range approximated based on the measurements

performed during the experiment. In addition, EnergyPlus-generated cooling energy savings are

at least five times lower than the earlier findings from different U.S. research studies summarized

(Medina, 2012).

Cooling Energy Consumption / Test Unit

Without Radiation Control Technology

With Radiation Control Technology

Annual Cooling Energy Savings

1A (Baseline) 4333 kWh N/A N/A 1B (IRCC) 4326 kWh 4308 kWh 0.43% 2A (RB) 4703 kWh 4703 kWh 1.27% 2B (IRCC) 4922 kWh 4922 kWh 0.83%

Table 1. Predicted Energy cooling Consumption and Associated Savings from Radiation Control Technologies

The modeling results in Table 1 show that cooling energy is saved with the use of radiation

control technologies. The table lists and compares the modeled annual cooling energy

consumption of each house before and after implementing the radiation control technology. The

“Annual Cooling Energy Savings” for each house in Table 1 is calculated by comparing cooling

energy consumption in “Without Radiation Control Technology” and “With Radiation Control

Technology” columns. The RB shows the highest savings from its lower emissivity and the air

gap created between the roof and the space between the two layers of the foil. Comparing the

two duplexes with IRCC installed, the results show that the attic geometry and the amount of

roof coated with the low-e material make a difference in the performance.

Parametric studies of the convection coefficients, air exchange rate, set point temperature and

changes of the heat transfer rate through exterior walls and windows were performed in the

process of calibrating the model. Convection coefficients were specified from the ASHRAE

Handbook of Fundamentals. The values for horizontal down flow were used and a slight

improvement was expected from the values EnergyPlus calculated. This was expected since all

roof surfaces were at angles above horizontal. The air exchange rate made little difference in the

output of the model. This is likely because the exchange of air has a greater effect on the zone air

temperature than the surfaces temperatures, which govern radiant heat flow. Heat flow through

the exterior walls and windows had the greatest effect on the performance of the RB since heat

flow through the ceiling into the conditioned zone was proportionally less than the other surfaces

even though the heat flux through the ceiling remained the same.

It should be noted that the above savings have been estimated based on EnergyPlus modeling

parameters close to existing test house conditions. Depending on HVAC configuration, air

leakage rate, and building envelope parameters, the savings can increase close to twice the values

reported in Table 1. However, they will still be significantly lower than the results of earlier

experimental studies (Medina 2012). This is one of the most surprising findings from this study.

Recorded temperature and energy consumption data showed that one test unit was notably

different during the winter 2011, which was evidence of an extensive use of an electric heater.

The use of a supplemental heater was unexpected, and an electric heater was not instrumented.

An additional challenge in modeling was caused by the fact that the tenant changed in one of the

test units just before the start of measurements. Also, the inability to control setpoint differences

between the units made direct comparisons difficult (since test units were occupied by tenants

with various space conditioning preferences).

One of the conclusions from this study is the fact that it is extremely difficult to perform a

detailed performance analysis of a single building enclosure system like RB using field test data

from occupied vintage housing units. It is still unclear how well energy consumption predictions

of the building systems containing RBs can be simulated using simplified Energy Plus roof/attic

algorithm. The team recommends that in the future, a calorimetric type of a field experiment

(with precisely measured attic air leakage rates, level of attic insulation, and with simple attic

structural components) be performed in order to further validate EnergyPlus predictions in cases

of RB applications. This experiment will most likely require test huts or similar small and well-

instrumented test structures.

5 Conclusion The major goal of this project was to evaluate the energy effects of RBs and IRCCs in field

conditions. Major focus was paid on thermal and energy performance of the attics. During the

cooling season 2012, thermal performance measurements were performed on four housing units

in two residential duplexes. Three test attics were modified using RBs and IRCCs and tested,

while the fourth attic stayed unmodified and was used as a base case for comparisons.

Thermal performance data was collected for each attic. Field experiments took place in Austin,

Texas. All test houses were occupied during the testing by residents. The original plan was to

keep internal load schedules in the test units as close as possible; however, each of these tenants

had different occupation habits and thermal comfort preferences. Also, individual schedules for

operating the AC were different. In addition, the test houses had attics with varying structural

components, thicknesses and types of attic floor insulation, and levels of attic ventilation. All of

the above factors made direct comparison of the whole building energy consumption impossible.

Measured test data confirmed earlier results from other research organizations. An average 3.4°C

reduction of the average attic air temperature caused by the multi-foil RB was recorded during

mid-summer 2012, and can be equivalent to over 34% reduction of the attic-generated cooling

loads—assuming approximate steady-state heat transfer, average attic temperature of 35°C and

internal air temperature of 25°C. This is consistent with a range of the cooling load reductions

reported by Medina (2012). It has been also reported that for U.S. residential houses, roofs and

attics generate an average 12% cooling energy contributions (Huang, et al., 1996). In that light,

potential whole building cooling energy savings may reach about 4%, depending on other

building parameters and HVAC system efficiency.

Similarly, a 2.4°C reduction of the average attic air temperature caused by the IRCC technology

can be an equivalent to over 24% reduction of the attic-generated cooling loads, and potential

whole building cooling energy savings for this technology may be close to 3%. Based on the

recorded data, energy performance of two IRCC systems seems to be very similar considering

the small differences in construction of both test attics in Duplexes 1 and 2.

An EnergyPlus computer model calibrated against both gas and electricity historical bill data for

the periods of pre-retrofit and post-retrofit was used for prediction of potential cooling energy

savings. The results from EnergyPlus were well below the range of cooling energy savings

approximated based on the measurements performed during this experiment. In addition,

EnergyPlus-generated cooling energy savings were at least four to five times lower than results

available from the earlier studies (Medina, 2012). It is recommends more work on validation of

whole building computer models using attic algorithms with RB.

Reference ASHRAE. 2007. ANSI/ASHRAE 90.2. Energy Efficient Design of Low-Rise Residential

Buildings. Atlanta: American Society of Heating, Refrigerating and Air Conditioning Engineers,

Inc.;

ASTM Standard C1313. 2010. Standard Specification for Sheet Radiant Barriers for Building

Construction Applications. ASTM International, West Conshohocken, PA, DOI:

10.1520/C1313_C1313M-10;

ASTM Standard C1340. 2010. Standard Practice for Estimation of Heat Gain or Loss Through

Ceilings Under Attics Containing Radiant Barriers by Use of a Computer Program. ASTM

International, West Conshohocken, PA, DOI: 10.1520/C1340_C1340M-10;

ASTM Standard C1321. 2009. Standard Practice for Installation and Use of Interior Radiation

Control Coating Systems (IRCCS) in Building Construction. ASTM International, West

Conshohocken, PA, DOI: 10.1520/C1321-09;

Atherton, S.A. 2011. Residential Attic Ventilation in A Hot And Humid Climate: Effects Of

Increased Ventilation On Thermal Performance And Moisture Control. Master Thesis,

University of Tennessee, Knoxville;

Handcock, E. Norton, P. and Hendron, B. 2002. Building America System Performance Test

Practices: Part 2, Air Exchange Measurements. NREL Technical Report NREL/TP-550-30270;

Accessed at: http://www.nrel.gov/docs/fy02osti/30270.pdf.

Huang, J. Hanford J. and Yang F. 1996. Residential Heating and Cooling Loads Component

Analysis. LBNL Technical Report LBNL-44636; Accessed at:

http://simulationresearch.lbl.gov/dirpubs/44636.pdf.

Medina, M. 2012. A Comprehensive Review of Radiant Barrier Research Including Laboratory

and Field Experiments. ASHRAE Transactions. Vol. 118 Issue 1, p400;

Parker, D. S. 2005. Literature Review of the Impact and Need for Attic Ventilation in Florida

Homes. Florida Solar Energy Center Report FSEC-CR-1496-05; Accessed at:

http://www.fsec.ucf.edu/en/publications/pdf/FSEC-CR-1496-05.pdf

Parker, D. S. 2003. Research highlights from a large scale residential monitoring study in a hot

climate . Energy and Buildings. 35 (9), pp. 863-87;

Parker, D. and Sherwin, J. 1998. Comparative Summer Attic Thermal Performance of Six Roof

Constructions. Proceedings of the 1998 ASHRAE Annual Meeting. Toronto, Canada;

RIMA International. 2002. Understanding and Using Reflective Insulation, Radiant Barriers and

Radiation Control Coating . RIMA International, Second Edition; Accessed at:

http://www.rimainternational.org/pdf/handbook.pdf.

RS Means. 2011. Building Construction Cost Data. Norwell, USA;

Stovall, T. Yarbrough D. W. Shrestha S. Pearson T. and Arimilli R. V. 2010. Analysis in Support

of the Radiant Barrier Fact Sheet 2010 Update. Proceedings of Thermal Performance of Exterior

Envelopes of Whole Buildings XI , Florida. Accessed at:

http://web.ornl.gov/sci/buildings/2012/B11%20papers/141_Stovall.pdf

Texas Gas Service, Home Energy Calculator; Available at:

http://www.texasgasservice.com/en/SaveEnergyAndMoney/HomeEnergyCalculator.aspx

Wallace, L.A. Emmerich, S.J. and Howard-Reed, C. 2002. Continuous measurement of air

change rates in an occupied house for 1 year: the effect of temperature, wind, fan and windows.

Journal of Exposure Analysis and Environmental Epidemiology (12) pp. 296- 306.

National Institute of Building Sciences Provider #G168

BEST4 Conference Nibsbest4 • April 13-15, 2015

Credit(s) earned on completion of this course will be reported to AIA CES for AIA members. Certificates of Completion for both AIA members and non-AIA members are available upon request. This course is registered with AIA CES for continuing professional education. As such, it does not include content that may be deemed or construed to be an approval or endorsement by the AIA of any material of construction or any method or manner of handling, using, distributing, or dealing in any material or product. ___________________________________________ Questions related to specific materials, methods, and services will be addressed at the conclusion of this presentation.

Participants will : 1. Learn how to link the performance of individual building enclosure components in a holistic framework to achieve high-performance buildings. 2. Explore, through built case studies, how building envelope design determines overall energy conservation and sustainability capabilities 3. Learn innovative practices for avoiding heat loss as well as moisture and air infiltration in enclosure design for healthy new and existing buildings. 4. Understand the role of building enclosure commission- ing in the design, construction, and operation and maintenance of commercial facilities.

Learning Objectives

Field Thermal Performance of EIFS Using Vacuum Insulation Panels

Ali Fallahi, Ph.D. Fraunhofer CSE, Boston, MA

Jan Kosny, Ph.D. Fraunhofer CSE, Boston, MA

Nitin Shukla, Ph.D. Fraunhofer CSE, Boston, MA

Tony Fontanini, Ph.D. Candidate Fraunhofer CSE, Boston, MA

Lawrence D. Carbary, F.ASTM Dow Corning Corporation, Midland, MI

Roland Serino, P.E. Dryvit Systems Inc., West Warwick, RI

Sometimes, it is impractical or costly to install thick insulation due to space loss or additional modifications and detailing

Wall Deep Energy Retrofit

Requires roof extension/detailing

Requires window relocation/detailing

1

Thinner solutions for a given target thermal resistance VIP offers one of the highest thermal resistances of any

insulation technology

Benefits of Vacuum Insulation Panels (VIPs)

VIPs are 4 to 8 times thinner than convention insulations for the same target R-vale

2

IEA/ECBCS Annex39, VIP Subtask A

Vacuum Insulation Panel (VIP) consists of a nanoporous core that is evacuated and enclosed in a container that is highly impervious to air and water vapor transmission

Thermal conductivity as low as 0.004 W/m-K

multi-layer envelope film

pressed silica core with opacifier

Core-bag

http://www.ecbcs.org/docs/Annex_39_Report_Subtask-A.pdf

Vacuum Insulation Panel (VIP) Technology

3

RCGST λλλλλ +++=

Heat Transport in Conventional Insulations

IEA/ECBCS Annex39, VIP Subtask A

λG, λC

λR λS

Conduction Convection Radiation

4

Macroporous media used at atmospheric pressure

Mineral fiber and polymeric foams

Porous structure implies only limited number of channel for heat conduction

Gas conduction dominates the thermal conductivity

Heat Transport in VIPs

Lower gas pressure

Reduce pore size

MFP for air molecules is ~70 nm VIP core is made of a nanoporous material Typical pore size is ~10-100 nm Core is surrounded by a film and evacuated to lower

pressures

5

Vacuum Insulation Panels – Thermal Bridging

IEA/ECBCS Annex39, VIP Subtask A

VIP

VIP

VIP

VIP

VIP Thermal Bridging due to Protective Foam- 3D Thermal Model showing heat loss through foam layer

Foil

VIP Thermal Bridging due to Metallic Foil

6

Project Objective: To assess the field performance of the VIP-based EIFS technology in building retrofit applications.

Subtasks: • Detailed instrumentation for field testing to

measure temperature and moisture gradient in the walls

• Performing small scale lab testing of material samples as an input for numerical modeling and hygrothermal analysis

• Evaluate and analyze the collected field test data

• Numerical analysis and energy simulations using EnergyPlus and WUFI to validate and extend performance evaluation

Before Retrofit

After Retrofit

Project Introduction

7

VIP

EPS

Wall Retrofit -Vacuum Insulation Panels

Two major U.S. building envelope manufacturers carried out the retrofit stage of this study, including VIP manufacturing, foam encapsulating, designing, and installation.

8

In addition, a major U.S. window manufacturer donated its high efficient R-5 windows to be installed as a part of the retrofit stage.

Wall Retrofit – High Performance Windows

Existing poor performing windows

High performing R-5 windows after retrofit

9

1

During Wall Retrofit

2

3 4 13 10

Before Retrofit After Retrofit

Wall Retrofit

11

Wall Sensor Layout

12

Inside View

West Elevation Sensors

Outside View

Wall Sensor Layout

13

Inside View

Field Test Collected Data

14

TIME Without Windows Sealed-off

With Windows Sealed-off

Difference

Pre-Retrofit 11.7 ACH @ 50 Pa 10.90 ACH @ 50 Pa 6%

Post- Retrofit 1 8.15 ACH @ 50 Pa 8.13 ACH @ 50 Pa 0.2% 30% Improved

Blower Door Test

15

Thermography Results

Before

After

16

Energy Model

Gas Consumption Calibration

Base-Case EnergyPlus model developed using actual weather conditions at the site of test building in Brunswick, ME

Base-Case model calibrated against historical utility bills both from pre-retrofit and post-retrofit stages. EnergyPlus Model of Test Building

17

Energy Model Configurations

Modeled EnergyPlus Configurations

Case A -Existing walls, R-3.7 -Existing windows, R-1 -Pre-retrofit Infiltration: 11.7 ACH

Case B -Existing walls, R-3.7 -High Efficient R-5 windows -Pre-retrofit Infiltration: 11.7 ACH

Case C -VIP walls, R-48.5 -Existing windows, R-1 -Pre-retrofit Infiltration: 8.1 ACH

Case D -VIP walls, R-48.5 -High Efficient R-5 windows -Pre-retrofit Infiltration: 8.1 ACH

Case E -VIP walls, R-48.5 -High Efficient R-5 windows -Pre-retrofit Infiltration: 3.5 ACH

18

Energy Model Results

0%

20%

40%

60%

80%

Case B Case C Case D Case E

Savi

ngs (

%)

Annual Gas Consumption Savings

Monthly Gas Consumption

19

Hygrothermal Model

Installed Moisture Pins

Modeled Wall Section in WUFI

20

Hygrothermal Model

MC% of West Wall and North Wall below 20% Safety Threshold for Wood Rotting

Typical Brunswick Climate

21

Thermographic images taken before and after the renovation clearly show improvement in heat losses after renovation with VIPs.

Blower door tests performed before and after the renovation show a 30% air tightness improvement after renovation with VIPs.

Based on WUFI modeling results and the measurements, the risk of moisture accumulation in the VIP-based EIFS of the retrofitted wall in the Brunswick building is low and remain below 12% and don’t reach the critical threshold for wood deteriorations.

Based on developed energy model of the Brunswick building, annual gas consumption savings of the building retrofitted with high efficient windows is 16% while in combination with other efficiency measures such as VIP-base EIFS and improved airtight building the savings of 48% and 70% are achievable.

Conclusion

25

Any Questions?